Children from low-income families are at heightened risk for developing language delays, poor social skills, and behavior problems. Early identification of the potential factors that may contribute to children's behavior problems and poor social skills is critical so that effective strategies for prevention and intervention can be provided. Few studies have explored the potential moderating effects of maternal depressive symptoms, home, and classroom environments on the relationship between child language and behavior. The proposed study represents the first attempt to use a longitudinal design and cutting edge observational methods to address this urgent public health problem. The Specific Aims are: (1) Determine whether children's early language abilities predict later trajectories of behavioral problems and poor social skills and whether the strength of this prediction differs for monolingual English-speaking children versus bilingual Hispanic children. (2) Determine whether English and Spanish language abilities predict behavior problems and social skills differentially for bilingual Hispanic children. (3) Determine the degree to which levels of home environment, classroom quality, and maternal depressive symptoms moderate the longitudinal relationships between children's language ability and their later behavior problems and social skills. Methods: Approximately 200 3- to 4-year-old children who attend Head Start and their mothers will participate in the study. English-speaking children's language skills will be assessed at T1, T2 (12 months later), and T3 (6 months later). Spanish-only speaking children will be assessed in Spanish on a standardized language test and a language screener at T1, and will be assessed in both Spanish and English at T2 and T3. Measures of children's behavior problems and social skills will be collected from teacher and parent reports at three time points. Each of the 200 children will be observed in their classrooms for 40 minutes at three time points with an innovative and comprehensive direct observational system. Mothers will complete a maternal depression scale and Home Screening Questionnaire at T1. Classroom quality will be assessed with Classroom Assessment Scoring System and the Classroom Code for Interactive Recording of Children's Learning Environment. Structural equation models will be used to demonstrate whether changes in children's early language abilities are linked strongly to behavioral trajectories over time and whether maternal depression and home and classroom qualities moderate the relationship between children's language and behavior. Significance: The successful completion of the aims will advance the theories regarding relationships of early language abilities to later behavior problems and to later social skills by providing new insights into protective and risk factors that may moderate these important relations. These theories and their important practical implications are utilized by special educators, speech/language pathologists, developmental psychologists, clinical psychologists, and psychiatrists.